二番目の議論は語順である。Everettによれば，ピダハン語の基本語順は SOVであり，他の言語で埋め込み節とみなされるものが目的語になる時も，
その埋め込み節は目的語であるから動詞の前に来るはずであるが，実際はそ の埋め込み節は動詞の後に来る。このことは，この動詞の後に置かれる節 は動詞の目的語ではなく，二つの節が並置されていると解釈すれば説明が つくということである。これに対して，Nevins et al. (2009a)は，基本語順が SOVである言語で目的語が節の場合に，その目的語の節が動詞の後にくるの はよく見られる現象である（Nevins et al. (2009a)はヒンディー語，ドイツ語，
Wappo語の例を挙げている）と反論している。
Everett’s principal argument for parataxis concerns word order. When the direct object is a nominal, Pirahã typically shows OV order. By contrast, examples like 13–16 must be taken to show VO order, if the -sai clauses are true objects of the finite verb. It might look at first glance, therefore, as though we have a simplicity argument in favor of the thesis that Pirahã lacks clausal embedding. In fact, however, the argument points the other way. As documented crosslinguistically by Dryer (1980), it is quite common to find clausal complements following a main verb in languages that are otherwise generally or uniformly OV. This is the case in Hindi, German, and Wappo, as in 19–21. (p. 372)
ピダハン語で，節が動詞の目的語になる時に動詞の後に置かれるのはIEPの ためではなく，世界の言語で普通に見られる現象であるとNevins et al. (2009a)
は説明する。
The word-order pattern seen in Pirahã is thus not at all unusual for a language with embedded clauses. The postverbal position of clausal arguments presents no puzzles that require special explanation in terms of a general ban on embedding or the IEP. In displaying VO word order where the object is a clause, Pirahã makes a typologically common choice from the menu of possibilities available to the languages of the world—the choice of a postverbal clausal complement in an otherwise OV language. This is a choice made by languages associated with diverse cultures and by languages with a varied range of other linguistic properties. (p. 373)
上の引用文中で，“As documented crosslinguistically by Dryer (1980)”とある が，Nevins et al. (2009a)は注24で，そのDryer (1980)の一般化を引用している。
Dryer (1980:128) offers the following generalization (his ‘final-over-internal-position hypothesis’): ‘Whenever sentential NPs of the same grammatical relation differ in their relative tendencies to occur in clause-final position as opposed to clause-internal position, the difference will be that sentential NPs will exhibit a greater tendency than simple NPs to occur in clause-final position rather than clause-internal position’. (p. 372)
なぜこのようなことが起こるのかについて，Nevins et al. (2009a)は，言語 処理の観点から，OV言語でも重い要素は動詞の後に置かれることがあると いうHawkins (1990, 1994)の仮説を紹介している。
. . . Hawkins (1990, 1994) suggests that considerations of sentence processing
guide the positioning of structurally complex phrases such as clauses. A hypothesis of this last type leads one to expect, in contrast to the purely syntactic proposals, that VO might be a general option in languages like Pirahã, dispreferred but chosen when the object exceeds some threshold of heaviness or complexity. (p. 374)
そして，ピダハン語でも（節ではなく）名詞でも重いものが動詞の後に置 かれることがあることをEverett自身の例を引用して紹介している。したがっ て，名詞目的語は動詞の前に来なくてはならないというEverettの重要な前提 が成立するかどうかは不明であり，複雑な目的語や重い目的語が動詞の後に くるという選択があるのであれば，-saiの付加された補文が動詞の後に来て もおかしくはないのである。
The extant data on Pirahã provide some evidence for the last of these possibilities [Hawkinsの仮説(S. N.)]. Although SOV is the dominant order in Pirahã, Everett’s earlier work offers a number of examples of SVO order in which the object is not a clause, but a heavy NP, as well as other examples where the postverbal object might well have been focused (though actual prosodic data are not available). Some of these examples involve heavy objects with appositive modifiers (e.g. 22a,b).
(22) SVO in Pirahá (objects in italics)
a. ti xoba-i-sog-abagaí hiaitíihí ti xahaigí I see-EP-DESID-FRUST.INIT Pirahã I brother
‘I want to see the Pirahã, who are my brothers.’ (HAL:212, ex. 55) b. tiobáhai koho-ái-hiab-a tomáti gihió-kasí píaii taí píaii child eat-ATEL-NEG-REM tomato bean.name also leaf also
‘(The) children do not eat tomatoes or beans or leaf(y vegetables).’
(HAL:226, ex. 107)
c. ti soxóá kap-í-hí baaí
I already shoot-PROX-COMP.CERT wild.pig
‘I already shot a wild pig.’ (HAL:295, ex. 361)
It is therefore not even clear that a NOMINAL object ‘should appear before the verb’—a key premise of Everett’s argument for parataxis over embedding. If postverbal position is a general option for complex or heavy objects in Pirahã, the postverbal position of a clausal -sai complement might be expected for reasons of heaviness or complexity alone. (p. 374)